What Goes into Our True Potato Seeds (TPS)
Everybody who produces true potato seed does it a little differently. Differences in the method of production might matter to you quite a bit. For example, there is a big difference between collecting berries that are largely self-pollinated and hand pollinating most of the flowers to ensure that seeds have greater genetic diversity. Our seeds are produced with the goal of maximizing diversity within particular phenotypes. This will help you to get to your goals faster. For that matter, it helps us to reach our goals faster, because a lot of what we grow comes from the same TPS that we sell.
We grow potatoes in phenotype blocks. A phenotype is the set of observable characteristics in a given variety: color, shape, flavor, etc. In this case, we’re mostly talking about skin and flesh color and shape. All red tetraploids are grown in a block, all blue tetraploids, all yellow flesh diploids, all elongated diploids, etc. This greatly increases the odds that insect pollinations will occur within the blocks, helping to preserve those phenotypes. Some varieties are planted in more than one block if they have a combination of characteristics appropriate to more than one. I hand pollinate daily within these blocks to ensure good berry set and minimize self-pollination. It takes about two hours each day to do the hand pollinations. First, I work though the block using a small paintbrush to apply the mixed pollen collected the previous day to each flower. Then I go back through the block and collect pollen, vibrating each flower with an electric toothbrush to collect pollen in a little cup for the following day. Some self-pollination is inevitable in this method, but by starting early in the day and pollinating before collecting pollen, we get much more diverse results. Some of our seed will also be insect pollinated from outside the block because our blocks are not spaced at a very great difference from one another. This is an intentional design. We have the space to plant the blocks at sufficient distance to prevent most cross pollination, but I want a small amount; I just don’t want the chaos of maximum pollination across blocks. I want that small amount of random cross pollination because that helps to ensure that there will be some combinations that I would never intentionally make. It is an insurance policy taken out against ignorance.
With one exception, the Kitchen Sink Mix, we only include seeds in our mixes that come from named varieties, breeding lines, and breeding line candidates. Those varieties make up only about 20% of what we grow. In any given year, we grow anywhere from about 500 to 3000 different varieties, most of which are seedlings that exist for only one year. We typically carry over somewhere between 200 to 300 varieties from year to year. It is those varieties that we carry over that primarily contribute to our seed mixes. Named varieties are those that either we have named and released or that have been by somebody else. They are the elite varieties. Breeding lines are those that we keep for their valuable traits, even though they might not quite make the cut for release. Breeding line candidates are the top performers in a given year that we haven’t grown before. So, our mixes contain seeds from roughly the best 20% of what we grow in any given year.
Why don’t we include everything? Because performance of the remaining 80% is uneven. I know that if you grow out a packet of one of our mixes, you are statistically likely to get some very good potatoes. That is not true with the remainder. You very well might get something great, but the odds are much lower. You will probably get some very interesting potatoes from the remainder, but they will have problems like poor disease resistance, late maturity, short day tuberization, or poor dormancy that will be serious problems for most growers. Because we have so much extra seed and because some people are interested in the high diversity, we introduced the Kitchen Sink Mix in 2016. It is basically a snapshot of everything that we grew in a given year, although you would have to grow several thousand seeds to get seedlings from all the possible parents.
Once berries start to ripen, we collect them by variety. When the tuber harvest comes, we decide which varieties will be carried over into the next year. The berries from those plants are then grouped by phenotype and processed. This is another stage of the process that mixes in some extra diversity. For example, let’s say that I am putting together the red flesh phenotype mix for the year. First, I add all the berries from named varieties and breeding lines with red flesh. These will have a majority of red progeny because they have been pollinated within the red block. Then, I add berries from breeding line candidates that have red flesh. These will have been grown in a seedling block, pollinated by other seedlings that have not yet been characterized. Typically, the berries from the named varieties and breeding lines will outnumber those produced by the seedlings by three or four to one. So, the new blood is likely to contribute more off phenotypes, but it is a smaller amount of seed.
I am focused on producing mixes rather than seed from individual breeding lines. This is because it is much easier for us to process and sell mixes and because I believe that the results are better for you in most cases. In any given year, I have grown out seed previously from the named varieties and breeding lines in a block, so I know that the seeds will produce a reasonable amount of the desired phenotype. Part of the mix will be from seedlings, in which case I have never grown out the seed to see the progeny. I know that they are statistically likely to produce at least a little of the desired phenotype in most cases, but some will not. That is why these seeds make up a smaller part of the mix.
Our packets contain enough seeds to make it very likely that you will get the phenotype advertised. Many people have a hard time understanding this part, so here is an example. In 2015, I grew out a packet of our Blue Flesh Tetraploid Mix. Packets contain 100 seeds. My goal for every packet of a particular phenotype mix is that you will get at least 20% seedlings that match the advertised phenotype. Remember, we’re breeding here and you should expect to throw away a lot of the results.
From that packet, I got 81 seedlings that survived to produce tubers, of the following phenotypes:
- 34 blue skin with white or light yellow flesh
- 19 white to light yellow skin and flesh, some with blue eyes
- 17 blue skin with partially to fully blue flesh
- 7 red skin with white flesh
- 3 red skin with partially to fully red flesh
- 1 yellow diploid (probably a hanger-on somewhere in the processing gear)
I didn’t quite hit the 20% goal for the described phenotype on this packet, but I have grown out enough to know that we usually do a little better than that. I consider that a good result. If you grew out this packet of seeds, you would have 17 blue flesh potatoes to work with. You could save seed from them and do a little better in the next generation. That’s what we’re shooting for. It is also worth noting that many of the varieties that lacked blue flesh would have had the right genes for it, just buried by one or more distribution genes that prevented it from expressing. Many of the off types could still serve as parents to produce varieties with blue flesh. Some phenotypes are easier to achieve than others. I could produce a white flesh potato mix that would be on target at better than 90%, but almost nobody would want it.
Of course, this example supposes that you grew the whole packet. Many people don’t have the space to grow 100 potato plants. If you grow only 10, then there is a fair chance that you wouldn’t get a blue flesh variety. If you grow again the next year and continue until you have used up the packet, you should get there eventually. Some people find this frustrating, but there is no good way to produce a packet of seeds that will be all blue flesh. I could do it by including only self-pollinated seeds of a select few varieties, but they would make a poor foundation for further breeding. The goal here is to deliver the phenotype that you want along with sufficient genetic diversity that you won’t dead end.
We don’t offer controlled crosses (i.e. crosses produced by emasculating flowers and pollinating them with a specific variety). I do make a lot of controlled crosses and the results of those eventually end up in the phenotype blocks. I don’t sell controlled crosses simply because they are so labor intensive and also have a fairly low success rate, which can’t be determined until you grow out the seeds. I would have to charge a fortune for them and the results wouldn’t be as good as those from the mixes in most cases.
If that all seemed complicated, it is. I don’t have all the answers. I am constantly refining this process to get better results. There is no right or wrong way to do it, but some methods work better than others. I think the most important thing here is to have a good understanding of what goes into the seed so that you can also have reasonable expectations about what can come out of it.